A Facile Synthesis of a Polyhydroxylated 2-Azabicyclo[3.2.1]octane Damon D. Reed and Stephen C. Bergmeier* Department of Chemistry and Biochemistry, Clippinger Laboratories, Ohio UniVersity, Athens, Ohio 45701
[email protected] ReceiVed September 18, 2006
Synthetic or natural aza-sugars have shown promise as a therapeutic approach to a variety of disease states by acting as transition state mimics to sugar processing enzymes. Although the synthesis of functionalized bicyclo[3.2.1]octanes has been reported, the procedures are relatively long and low yielding. Herein, we report the facile synthesis of polyhydroxylated 2-azabicyclo[3.2.1]octane that can be selectively functionalized. The first naturally occurring sugar-mimic, i.e., aza-sugar or polyhydroxylated alkaloid, was isolated in 1966.1 Since then, numerous aza-sugars have been isolated and can be divided into five general classes: piperidines, pyrrolidines, indolizidines, pyrrolizidines, and nortropanes.2 Numerous polyhydroxylated alkaloids from these structural classes, synthetic and natural, have shown promise as anti-viral or anti-infective agents as well as in the treatment of diabetes.2 Aza-sugars inhibit enzymes involved in sugar processing by acting as transition state mimics.3 1-Deoxymannojirimycin (DNJ) is an example of a polyhydroxylated monocyclic alkaloid, aza-sugar, that acts as a glycosidase inhibitor.4 Currently, a derivative of DNJ, Nhydroxyethyl-DNJ, is marketed by Glaxo as Miglitol for the treatment of type II diabetes.5 As part of our program aimed at preparing analogues of the norditerpenoid alkaloid methyllycaconitine, we required an efficient synthesis of the aminosugar 3. Remarkably few examples of this ring system have been reported.6,7 While the synthesis of the trans-6,7-diol [3.2.1] * To whom correspondence should be addressed. Tel: 740-517-8462. Fax: 740-593-0148.
(1) Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G. W. J. Tetrahedron: Asymmetry 2000, 11, 1645-1680. (2) Pearson, M. S. M.; Mathe-Allainmat, M.; Fargeas, V.; Lebreton, J. Eur. J. Org. Chem. 2005, 2159-2191. (3) Lillelund, V. H.; Jensen, H. H.; Liang, X. F.; Bols, M. Chem. ReV. 2002, 102, 515-553. (4) Andersen, S. M.; Lundt, I.; Marcussen, J.; Yu, S. Carbohydr. Res. 2002, 337, 873-890. (5) Pearson, M. S. M.; Saad, R. O.; Dintinger, T.; Amri, H.; MatheAllainmat, M.; Lebreton, J. Bioorg. Med. Chem. Lett. 2006, 16, 32623267. (6) Fox, B. L.; Reboulet, J. E. J. Org. Chem. 1968, 33, 3639-3641. (7) Ong, H. H.; Anderson, V. B.; Wilker, J. C. J. Med. Chem. 1978, 21, 758-763.
azabicycle has been reported,8 the route is low yielding and does not allow for the introduction of the 4-hydroxyl group. An alternative synthesis prepared a lactone derivative of 3, which incorporated the 4-hydroxyl group, but the synthesis is quite long and low yielding.9 Therefore, we decided to investigate the selective functionalization of N-tosyl-2-azabicyclo[3.2.1]octa-3,6-diene (2) as a starting point for our synthesis of 3. Compound 2 is readily available,10,11 and the two double bonds can be chemoselectively differentiated to provide the target compound 3 (Scheme 1). The reaction of tosyl azide12 with norbornadiene, 1, yielded the desired N-tosyl-2-azabicyclo[3.2.1]octa-3,6-diene 2 in good yield.10,11 With compound 2 in hand, regioselective dihydroxylation of the unfunctionalized olefin was attempted with OsO4/ NMO under a variety of solvent and temperature conditions. Although the desired product 4 was isolated from OsO4/NMOmediated dihydroxylation, low yields, and difficulties with purification prompted the investigation of other dihydroxylation methodologies. Surprisingly, AD-mix R or β chemoselectively dihydroxylates the desired double bond to form the racemic exodiol 4 in excellent yield. The kinetic resolution of [2.2.1] bicyclic systems via AD-mix-mediated dihydroxylation provides little to no enantioselectivity;13 therefore, it is not surprising that there was no enantioselectivity in the dihydroxylation of 2 to form 4. Although previous research into the functionalization of [2.2.1] bicyclic systems has elucidated that the chemoselectivity between two olefins is typically the result of sterics and not electronics,14 further research needs to be done on this [3.2.1] azabicyclic system to determine which dictates the chemoselectivity observed in the dihydroxylation of 2 (Scheme 2). Yields associated with the protection of the exo-diol, as the acetonide 5, were dependent on the amount of time between purification of 4 and formation of the acetonide, 5. When diol 4 is stored as a neat liquid at room temperature under an argon atmosphere, two distinct compounds (mass 408 and 779) were observed. Both of these compounds subsequently decomposed to a number of unidentified products. The conversion of 4 to 5 immediately after isolation provided consistently high yields of 5. The next step was introduction of the hydroxyl at C-4. It has been noted that exocyclic hydroboration of related [2.2.1]heptene systems is controlled by the steric bulk of the bridgehead substituents.15 Our expectation was that 5 would follow this general trend. Treatment of 5 with 9-BBN followed by an oxidative workup yielded the desired compound, 6a, albeit in moderate yields. Clearly 5 appears to follow this general rule of [2.2.1] systems of preferential hydroboration in the exo position. In an effort to improve the yield, hydroboration with (8) Johansen, S. K.; Korno, H. T.; Lundt, I. Synthesis 1999, 171-177. (9) Gonzalez-Alvarez, C. M.; Quintero, L.; Santiesteban, F.; Fourrey, J. L. Eur. J. Org. Chem. 1999, 3085-3087. (10) Oehlschlager, A. C.; Zalkow, L. H. J. Org. Chem. 1965, 30, 42054211. (11) Rhee, H.; Yoon, D. O.; Jung, M. E. Nucleosides Nucleotides Nucleic Acids 2000, 19, 619-628. (12) Curphey, T. J. Org. Prep. Proc. Int. 1981, 13, 112-115. (13) Bakkeren, F.; Klunder, A. J. H.; Zwanenburg, B. Tetrahedron 1996, 52, 7901-7912. (14) Mayo, P.; Tam, W. Tetrahedron 2002, 58, 9527-9540. (15) Brown, H. C.; Knights, E. F.; Scouten, C. G. J. Am. Chem. Soc 1974, 96, 7765-7770. 10.1021/jo0619231 CCC: $37.00 © 2007 American Chemical Society
1024
J. Org. Chem. 2007, 72, 1024-1026
Published on Web 01/05/2007
SCHEME 1.
General Plan for Synthesis of 3
SCHEME 2.
Osmylation of Diene 2 FIGURE 1. Dihedral angles and observed coupling constants for diastereomeric alcohols.
SCHEME 3.
Hydroboration/oxidation of Enamine
FIGURE 2. Observed NOESY cross-peaks. SCHEME 4.
BH3‚THF was tested and subsequently provided a higher yield of 6a with a small amount of the endo isomer, 6b, which was easily separated via recrystallization (Scheme 3). The relative stereochemistry of bicyclo[2.2.1]heptane systems is often determined by the coupling constant between adjacent hydrogens as a result of the dihedral angle between the two hydrogens.14,16 This method appears to be applicable to the azabicyclo[3.2.1]octane system as well. As shown in Figure 1, the dihedral angle between 3Hexo and 4Hendo protons of 6a is calculated to be 77°.17 The observed coupling constant as expected is
